BananaVision: A banana foliage and banana diseases dataset for computer vision

BananaVision is envisioned to provide banana foliage image data in real-life farm conditions for various computer vision tasks including classification, detection, and segmentation. BananaVision images show whole plant structures and multiple views of a single plant within plantation environments. Currently, the dataset has 2343 images representing 700 plants, which covers healthy and diseased specimens, including those infected by Fusarium wilt, Moko, and Bunchy top.香蕉视觉数据集（BananaVision）旨在提供真实农田环境下的香蕉叶片图像数据，用于开展分类、检测与分割等多种计算机视觉任务。该数据集的图像展示了种植园环境中的完整植株结构，以及单株植株的多视角画面。目前，该数据集共包含2343张图像，对应700株植株，涵盖健康与染病样本，其中包括受镰刀菌枯萎病（Fusarium wilt）、莫科病（Moko）以及香蕉束顶病（Bunchy top）侵染的植株。

搜集汇总

背景与挑战

背景概述

BananaVision是一个专注于香蕉叶子和疾病的计算机视觉数据集，包含2343张图像，覆盖700株植物，用于分类、检测和分割任务。数据集包括健康和患病样本，特别是感染Fusarium wilt、Moko和Bunchy top等疾病的图像，旨在模拟真实农场条件以支持农业研究。